To read this content please select one of the options below:

(excl. tax) 30 days to view and download

Patent value evaluation based on Bayesian optimized XGBoost model

Panjun Gao, Yong Qi, Hongye Zhao, Xing Li

Kybernetes

ISSN: 0368-492X

Article publication date: 21 January 2025

27

Abstract

Purpose

The purpose of this study is to address the critical need for patent value evaluation within patent management, particularly in the context of the digital economy. Recognizing the importance of utilizing historical data, this research aims to uncover effective methodologies that enhance the appraisal of patent value, which is vital for informed decision-making in the management of scientific and technological advancements.

Design/methodology/approach

This study introduces a comprehensive evaluation model by analyzing various factors that influence patent value. An index system is constructed that integrates technical, economic and legal aspects to facilitate a nuanced assessment of patents. The methodological core of this research is the development of an XGBoost patent value appraisal model, which incorporates Bayesian optimization to refine the evaluation process. The model’s validity is tested through empirical analysis of patents in the rapidly evolving sector of cloud computing.

Findings

The empirical results demonstrate that the XGBoost model, strengthened by Bayesian optimization, outperforms traditional categorization techniques. The proposed model shows superior performance in terms of accuracy, precision, recall rate and operational feasibility. These findings indicate a significant improvement in the precision of patent potential and value assessments, leading to more reliable and actionable insights for patent management.

Originality/value

This study introduces a novel patent evaluation model that combines XGBoost with Bayesian optimization. XGBoost enhances performance by integrating weak learners, ideal for complex, nonlinear problems like patent valuation. Bayesian optimization refines hyperparameters efficiently using prior distributions and known results. Its practical implications for patent management and technology exploration are substantial, offering a new tool for strategic decision-making.

Keywords

Citation

Gao, P., Qi, Y., Zhao, H. and Li, X. (2025), "Patent value evaluation based on Bayesian optimized XGBoost model", Kybernetes, Vol. ahead-of-print No. ahead-of-print. https://doi.org/10.1108/K-05-2024-1264

Publisher

:

Emerald Publishing Limited

Copyright © 2025, Emerald Publishing Limited

Related articles